Models & Benchmarks

Models and IEC 61724#

Denowatts utilizes the IEC 61724-series standards for energy and power performance definitions, analysis, and reporting.

IEC 61724-3 (2016) defines “Model” as the physics-based model chain and simulation model, including parameters, used to calculate both predicted and expected PV generation from weather data. The model is also used to calculate expected energy during times of unavailability.

While the physics-based functions are universal in all Denowatts benchmarking, any unique input parameter update results in a new Model.

Denowatts utilizes two Model Types: The Investor’s Model and the Operator’s Model. Each Model Type fulfills a unique requirement:

• Investor’s Model: Provided to Denowatts by the Site Owner or Providers for reporting underwriting expectations. In the Deno Portal, the Investor’s Model is selected as the Owner’s (Expected) Model Type.
• Operator’s Model: Produced by Denowatts based on observed production for reporting optimal performance.

Models may be additionally characterized by their time-series weather data inputs, which are based on either a synthesized data set, often referred to as Typical Meteorological Year (TMY) or “historic”, or measured weather data.

Models	Synthesized/TMY	Measured
Investor	Predicted (IEC 61724-3)	Expected (IEC 61724-3)
Operator	Forecasted (Denowatts)	Learned (Denowatts)

The table columns are Time-Series Data Benchmarks.

Benchmarks#

Models utilize time-series input data to produce output data called Benchmarks. Depending on their type, Benchmarks are simulated either once, if using Synthesized input data, or in real-time, if using measured input data.

Predicted Energy Benchmark (Investor’s Model)#

The Predicted Energy benchmark is calculated using the owner-provided Model, which utilizes a TMY or similar P50 input weather dataset. The use of Predicted Energy in the Denowatts portal always refers to benchmarking against the Investor’s Model, as defined in IEC 61724-3.

Portal page: Predicted Energy Model

Expected Energy Benchmark (Investor’s Model)#

The Expected Energy benchmark is calculated using the owner-provided Model, which utilizes measured irradiance and cell temperature data. The use of Expected Energy and Expected Power in the Denowatts portal always refers to benchmarking against the Investor’s Model, as defined in IEC 61724-3.

Portal page: Expected

Forecasted Energy Benchmark (Operator’s Model)#

The Forecasted Energy benchmark is calculated using the Operator-produced Model, which utilizes a TMY or similar P50 input weather dataset. The use of Forecasted Energy in the Denowatts portal always refers to benchmarking against the Operator’s Model, as defined by Denowatts.

Learned Energy Benchmark (Operator’s Model)#

The Learned Energy benchmark is calculated using the Operator-produced Model, which utilizes measured irradiance and cell temperature data. The use of Learned Energy and Learned Power in the Denowatts portal always refers to benchmarking against the Operator’s Model, as defined by Denowatts.

Portal page: Learned

Related#

• Methodology & Specifications, section 4: model equations, the Learned Profile, Energy Accounting, and KPIs.
• Model Management: Simulations, Proxy Models, and Blocks.
• Energy Test introduction: Energy Accounting and the Predicted, Expected, and Learned reconciliation.
• Metrics and KPIs